If you're using them on a page file, or defragging them you're doing it wrong I would think.
In my experience, if a system with an SSD is set up correctly, the SSD does not get defragged automatically, ever (ie. no setting changes with regard to defrag config).
As for the page file, SSD or HDD, everyone's advice is to ensure that you have enough RAM so that the page file doesn't get used. HDD thrashing is not a good thing, neither is running with low RAM.
I don't know how often you guys change your HDDs but I wouldn't trust an SSD to last 10+ years, at least not yet.
Define "trust"...? I wouldn't have any appreciable level of confidence in any HDD to predict that it will last for more than ten years. Obviously I have seen some last that long, but again, it's common knowledge that the most common time for a HDD to fail is in the first five years.
As for SSDs I don't have that level of trust for them either because obviously I haven't been using them for that long, but the fact that they have no moving parts is a potential massive plus in their favour.
I still have HDDs from the early 90s - 2000s in use. The SSD seem like they're great when they're new but if they get old, they just won't last. It's true they'll handle physical shock better but if I want to have an older system just sit on the shelf and then boot it up some time later, there is a risk the data won't be there. I've had electrolytic capacitors fail before I've had HDDs fail. In fact, I plan to make a thread about the awesome quality of ECS's electrolytic capacitors in a motherboard that I THOUGHT would have steer clear of those doomed capacitors.
I think your argument applies just as well to any new technology. Better stick with those Pentium 3s, anything newer might be "unreliable". I've personally seen way more HDDs fail than motherboards, probably 10x more, and yes, I saw plenty of dead boards during the "capacitor plague".